Дисертації з теми "Diesel motor Mathematical models"

In this work we present a geometric model of motor cortex that generalizes an already existing model of visual cortex. The thesis opens by recalling the notions of fiber bundles, principal bundles, Lie groups and sub-Riemannian geometry. In particular, we enunciate Chow’s theorem which ensures that if the Hörmander condition holds, the space connectivity property is satisfied. Then we recall the visual cortex model proposed by Citti-Sarti, which describes the set of simple cells as a Lie group with sub-Riemannian metric. The original part of the thesis is the extension to the motor cortex. Based on neural data, collected by Georgopoulos, we study the set of motor cortical cells and we describe them as a principal bundle. The fiber contains the movement direction and shapes the hypercolumnar structure measured. Finally we determine the intrinsic coordinates of the motor cortex, studying the cellular response to the motor impulse.

Thesis (M.S.)--West Virginia University, 2006.
Title from document title page. Document formatted into pages; contains xiii, 85 p. : ill. (some col.). Vita. Includes abstract. Includes bibliographical references (p. 82-84).

The flagellar motor of Escherichia coli is a bidirectional rotary nano-motor, powered by a transmembrane influx of protons. The maximum speed of rotation is about 300 Hz. The motor rotates either counterclockwise (CCW) or clockwise (CW) and the rotation direction is controlled by a chemotactic protein, CheY-P. Despite extensive structural and functional studies, precise mechanisms regarding the torque generation and the directional switching processes remain unclear. In this work, a bottom-up strategy is proposed and followed to investigate this motor. This strategy, named as a multiscale modeling approach, integrates various publicly available experimental data and ‘state-of-the-art’ molecular modeling methods to build structural models for the two most important parts of the motor, the C ring and the stator. Starting from the primary sequences of the composition proteins of these two substructures, tertiary structures are predicted by means of comparative modeling or de novo prediction when the comparative modeling is not available. Quaternary structures of these proteins’s complexes are then predicted by data-driven protein-protein docking or multiscale molecular dynamics (MD) simulations. Finally, structural models of the C ring at CCW and CW rotational states are constructed by cryo-EM aided assembling methods (constraint search that is based on the multiscale modeling tools and under the constraint of the EM images). In the case of the stator, its composition proteins, MotA and MotB, are assembled by coarse-grained MD simulations. This is the first molecular model based atomistic details for the stator. A new molecular mechanism for rotational switching is proposed based on the structural models of the C ring and the stator. The two states of the C ring display significant differences in the interfaces among the self-assembled FliMs and the orientations of FliG C-terminal domain. Based on protein docking results, a binding site of CheY-P is identified on FliM which is close to the interfaces of FliMs for self-assembling. Thereby, I propose that the CheY-P binding interferes with the interactions between neighboring FliM proteins, and thus, induces ~65° rotation of the FliGc domain with respect to FliM. Subsequently, the interactions between the stator and FliGc domains are altered and the rotation direction is changed. Furthermore, a mechanism accounting for the directed rotation of the flagellar motor is proposed based on systematic MD studies on the dynamics of FliGc. It is found that the C-terminal subdomain of FliGc is capable of rotating by ~180° with respect to the N-terminal subdomain of FliGc. If this dynamics is considered in the framework of the whole C ring, the rotation of the C-terminal subdomain of FliGc exhibits an asymmetric feature. As a result, the C ring is decorated with asymmetric teeth on the outer periphery and hence similar with Feynman’s ratchet. The preference of the motor in CCW rotation or in CW rotation is then explained based on the Feynman’s ratchet model.
published_or_final_version
Chemistry
Doctoral
Doctor of Philosophy

Increasingly, large volumes of aircraft data are being recorded in an effort to adapt aircraft maintenance procedures from being time-based towards condition-based techniques. This study uses techniques of artificial intelligence and develops mathematical models to analyse this data to enable improvements to be made in aircraft management, affordability, availability, airworthiness and performance. In addition, it highlights the need to assess the integrity of data before further analysis and presents the benefits of fusing all relevant data sources together. The research effort consists of three separate investigations that were undertaken and brought together in order to provide a unified set of methods aimed at providing a safe, reliable, effective and efficient overall procedure. The three investigations are: 1. The management of helicopter Health Usage Monitoring System (HUMS) Condition Indicators (CIs) and their analysis, using a number of techniques, including adaptive thresholds and clustering. These techniques were applied to millions of CI values from Chinook HUMS data. 2. The identification of fixed-wing turbojet engine performance degradation, using anomaly detection techniques, applied to thousands of in-service engine runs from Tornado aircraft. 3. The creation of models to identify unusual aircraft behaviour, such as uncommanded flight control movements. Two Chinook helicopter systems were modelled and the models were applied to over seven hundred in-service flights. In each case, the existing techniques were directed toward a condition-based maintenance approach, giving improved detection and earlier warning of faults.

Position and vibration control of a flexible beam is studied analytically and in the laboratory. Two different motor types are compared as actuators throughout the thesis: a standard voltage controlled motor and a torque controlled motor. The experimental beam is controlled with a dc-servo motor at its base and is instrumented with strain gages and a potentiometer. The control law is a form of linear, direct-output feedback. State estimators augment the control law to provide rate information that is not available from the instrumentation. Accurate modeling of the system’s inherent damping characteristics is achieved by analyzing experimental data. Gains were iterated yielding minimum-gain norm and minimum-sensitivity norm solutions to meet imposed eigenvalue placement constraints. Results for the two solutions and the two systems are compared and contrasted. Experimental verification of analytical results is hampered by unmodeled system non-linearities. Several attempts at bypassing these obstacles are shown. Finally, conclusions and recommendations are made.
Master of Science
incomplete_metadata

Qualquer tarefa motora ativa se dá pela ativação de uma população de unidades motoras. Porém, devido a diversas dificuldades, tanto técnicas quanto éticas, não é possível medir a entrada sináptica dos motoneurônios em humanos. Por essas razões, o uso de modelos computacionais realistas de um núcleo de motoneurônios e as suas respectivas fibras musculares tem um importante papel no estudo do controle humano dos músculos. Entretanto, tais modelos são complexos e uma análise matemática é difícil. Neste texto é apresentada uma abordagem baseada em identificação de sistemas de um modelo realista de um núcleo de unidades motoras, com o objetivo de obter um modelo mais simples capaz de representar a transdução das entradas do núcleo de unidades motoras na força do músculo associado ao núcleo. A identificação de sistemas foi baseada em um algoritmo de mínimos quadrados ortogonal para achar um modelo NARMAX, sendo que a entrada considerada foi a condutância sináptica excitatória dendrítica total dos motoneurônios e a saída foi a força dos músculos produzida pelo núcleo de unidades motoras. O modelo identificado reproduziu o comportamento médio da saída do modelo computacional realista, mesmo para pares de sinal de entrada-saída não usados durante o processo de identificação do modelo, como sinais de força muscular modulados senoidalmente. Funções de resposta em frequência generalizada do núcleo de motoneurônios foram obtidas do modelo NARMAX, e levaram a que se inferisse que oscilações corticais na banda-beta (20 Hz) podem influenciar no controle da geração de força pela medula espinhal, comportamento do núcleo de motoneurônios até então desconhecido.
Any active motor task is accomplished by the activation of a motor unit population. However, due to many ethical and technical difficulties the synaptic input to the motoneurons cannot be measured in humans. For these reasons realistic computational models of a motoneuron nucleus and the innervated muscle fibers have an important role in the study of the human control of muscles. However such models are complex and their mathematical analysis is difficult. In this text a system identification approach of a realistic motor unit nucleus model is presented with the objective of obtaining a simpler model capable of representing the transduction of the motor unit nucleus inputs into the muscle force signal associated to that nucleus. The system identification was based on an orthogonal least squares algorithm to find a NARMAX model, the input being the net dendritic excitatory synaptic conductance of the motoneurons and the output being the muscle force signal produced by the motor unit nucleus. The identified model output reproduced the mean behavior of the output from the realistic computational model even for input-output signal pairs not used during the identification process, such as sinusoidally modulated output muscle force signals. Generalized frequency response functions of the motoneuron nucleus were obtained from the identified NARMAX model, and led to an inference that cortical oscillations in the beta band (20 Hz) can affect force control by the spinal cord, an unknown motoneuron nucleus behavior until now.

The area of powertrain dynamics has received considerable attention over a number of years. The recent introduction of more stringent emission requirements together with economic pressure has led to a particular focus on increasing powertrain efficiency. This has seen the incorporation of on-board, real-time measurements to predict system behaviour and engine condition. In this domain, accurate models for all powertrain components are important. One strategy to improve accuracy is to evaluate the assumptions made when deriving each model and then to address the simplifications that may introduce large errors. To this end, the aim of the work presented in this dissertation was to investigate the consequences of some of the more common assumptions and simplifications made in low frequency torsional powertrain models, and to propose improved models where appropriate. In particular, the effects of piston-tocylinder friction, crank/gudgeon pin offset, and the torsional behaviour of tyres were studied. Frequency and time domain models were used to investigate system behaviour and model predictions were compared with measurements on a small single cylinder engine. All time domain engine and powertrain models also include a variable inertia function for each reciprocating mechanism. It was found that piston-to-cylinder friction can increase the apparent inertia variation of a single reciprocating engine mechanism. This has implications for the nonlinear behaviour of engines and the drivetrains they are connected to. The effect of crank/gudgeon pin offset also modified the nonlinear behaviour of the mechanism. Though, for typical (small) gudgeon offset values these effects are small. However, for large offset values, achievable practically with crank offset, the modification to the nonlinear behaviour should not be ignored. The low frequency torsional damping properties of a small pneumatic tyre were found to be more accurately represented as hysteretic rather than viscous. Time domain modelling was then used to extend the results to a multi-cylinder engine powertrain and was achieved using the Time Domain Receptance (TDR) method. Various powertrain component TDRs were developed using Laplacians. Powertrain simulations showed that piston-to-cylinder friction can provide additional excitation to the system.

Orientador: Edson Bim
Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação
Made available in DSpace on 2018-07-21T11:58:56Z (GMT). No. of bitstreams: 1 Pereira_ValmirMachado_M.pdf: 3590571 bytes, checksum: 64d5729c458a96919ef78024257af6a4 (MD5) Previous issue date: 1996
Resumo: Esta dissertação tem o objetivo de determinar um modelo matemático para o motor de indução bifásico assimétrico com os enrolamentos de estator conectados em T e investigar algumas propriedades tais como aceleração livre, característica estática conjugado-velocidade e as correntes em regime permanente. Este motor é projetado para ser alimentado a partir de duas fases e do neutro, e requer um valor particular para a relação de espiras entre os enrolamentos do estator. Inicialmente, uma transformação de variáveis é realizada com a finalidade de encontrar um motor equivalente que não possua nenhum ponto condutivo comum entre os enrolamentos e que apresente a mesma fm.m do motor original. Fazendo uso do conceito de circuitos acoplados são obtidas as equações básicas. A seguir, essas variáveis são escritas no sistema de referência estacionário e um circuito equivalente é obtido. Utilizando esse circuito equivalente, são apresentados resultados de simulação computacional e resultados de testes experimentais.A comparação dos mesmos exibiram uma boa concordância,mostrando que o modelo matemático é adequado
Abstract: This dissertation has the aim of determining a mathematical model for asymmetrical two-phase induction motor with T-connected stator windings and to investigate some properties such as free acceleration, steady-state torque-speed characteristic and steady-state currents. This motor is designed to operate from supply of two phases and neutral, and a particular value for the stator tums ratio is required. First, a transformation of variables is accomplishedto obtain an equivalent motor that do not have electric connection and that presents the same m.m.f. of original motor. Making use of the concept couple circuits, the basic equations are derived. Second, these variables are written in the stationary referencefTameand an equivalentcircuit is obtained. By using this equivalent circuit, computer simulationis presented and compared with experimental results. The tests exhibited a good agreement, showing that the mathematical model is adequate
Mestrado
Mestre em Engenharia Elétrica

Cable-actuated manipulators (or 'cable robots') constitute a relatively new classification of robots which use motors, located at fixed remote locations, to manipulate an end-effector by extending or retracting cables. These manipulators possess a number of unique properties which make them proficient with tasks involving high payloads, large workspaces, and dangerous or contaminated environments. However, a number of challenges exist which have limited the mainstream emergence of cable robots. This thesis addresses two of the most important of these issues-- workspace analysis and disturbance compensation. Workspace issues are particularly important, as many large-scale applications require the end-effector to operate in regions of a particular shape, and to exert certain minimum forces throughout those regions. The 'Force-Feasible Workspace' represents the set of end-effector positions, for a given robot design, for which the robot can exert a set of required forces on its environment. This can be considered as the robot's 'usable' workspace, and an analysis of this workspace shape for point-mass cable robots is therefore presented to facilitate optimal cable robot design. Numerical simulation results are also presented to validate the analytical results, and to aid visualization of certain complex workspace shapes. Some cable robot applications may require mounting motors to moving bases (i.e. mobile robots) or other surfaces which are subject to disturbances (i.e. helicopters or crane arms). Such disturbances can propagate to the end-effector and cause undesired motion, so the rejection of motor mount disturbances is also of interest. This thesis presents a strategy for measuring these disturbances and compensating for them. General approaches and implementation issues are explored qualitatively with a simple one-degree-of-freedom prototype (including a strategy for mitigating accelerometer drift), and quantitative simulation results are presented as a proof of concept.

This work presents an investigation into the parameter estimation of suspension components and the vertical motions of wheeled vehicles from experimental data. The estimation problems considered were for suspension dampers, a single wheel station and a full vehicle. Using conventional methods (gradient-based (GB), Downhill Simplex (DS)) and stochastic methods (Genetic Algorithm (GA) and Differential Evolution (DE)), three major problems were encountered. These were concerned with the ability and consistency of finding the global optimum solution, time consumption in the estimation process, and the difficulties in setting the algorithm's control parameters. To overcome these problems, a new technique named the discrete variable Hybrid Differential Evolution (dvHDE) method is presented. The new dvHDE method employs an integer-encoding technique and treats all parameters involved in the same unified way as discrete variables, and embeds two mechanisms that can be used to deal with convergence difficulties and reduce the time consumed in the optimisation process. The dvHDE algorithm has been validated against the conventional GB, DS and DE techniques and was shown to be more efficient and effective in all but the simplest cases. Its robustness was demonstrated by its application to a number of vehicle related problems of increasing complexity. These include case studies involving parameter estimation using experimental data from tests on automotive dampers, a single wheel station and a full vehicle. The investigation has shown that the proposed dvHDE method, when compared to the other methods, was the best for finding the global optimum solutions in a short time. It is recommended for nonlinear vehicle suspension models and other similar systems.

This work presents an investigation into the parameter estimation of suspension components and the vertical motions of wheeled vehicles from experimental data. The estimation problems considered were for suspension dampers, a single wheel station and a full vehicle. Using conventional methods (gradient-based (GB), Downhill Simplex (DS)) and stochastic methods (Genetic Algorithm (GA) and Differential Evolution (DE)), three major problems were encountered. These were concerned with the ability and consistency of finding the global optimum solution, time consumption in the estimation process, and the difficulties in setting the algorithm's control parameters. To overcome these problems, a new technique named the discrete variable Hybrid Differential Evolution (dvHDE) method is presented. The new dvHDE method employs an integer-encoding technique and treats all parameters involved in the same unified way as discrete variables, and embeds two mechanisms that can be used to deal with convergence difficulties and reduce the time consumed in the optimisation process. The dvHDE algorithm has been validated against the conventional GB, DS and DE techniques and was shown to be more efficient and effective in all but the simplest cases. Its robustness was demonstrated by its application to a number of vehicle related problems of increasing complexity. These include case studies involving parameter estimation using experimental data from tests on automotive dampers, a single wheel station and a full vehicle. The investigation has shown that the proposed dvHDE method, when compared to the other methods, was the best for finding the global optimum solutions in a short time. It is recommended for nonlinear vehicle suspension models and other similar systems.

Orientador: Ernesto Ruppert Filho
Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e Computação
Made available in DSpace on 2018-08-11T20:28:16Z (GMT). No. of bitstreams: 1 LuqueCarcasi_DiodomiroBaldomero_M.pdf: 943595 bytes, checksum: c20084d6cea7e72315ca4292797e783f (MD5) Previous issue date: 2005
Resumo: Este trabalho apresenta um estudo sobre o comportamento de motores de indução trifásico em regime permanente. Discute-se a validade da utilização do circuito elétrico equivalente do motor nas condições nominais e fora das condições nominais de operação. Ensaios em vazio e com rotor bloqueado são executados em um motor de indução trifásico para a determinação dos parâmetros do circuito equivalente. O ensaio realizado em condições de carga variável dentro de um intervalo usual de operação do motor, em termos de escorregamento, mostra que o circuito equivalente não representa o comportamento da máquina quando esta trabalha fora do ponto nominal de operação, porque os parâmetros não mais o representam bem naquelas situações de operação. As grandezas obtidas nos ensaios em carga variável são adequadamente condicionadas e utilizadas no treinamento de uma rede neural artificial Multi-Layer Perceptron (MLP). Para a validação dessa rede são comparados os resultados obtidos no modelo neural com os testes experimentais do motor de indução trifásico.
Abstract: This thesis presents a study about the steady state behavior of the three-phase induction motor. The validity of the use of the electric equivalent circuit is discussed when it is utilized on the rated conditions and also on any other operating conditions. No-load and the blocked rotor tests are performed on a three-phase induction motor for the determination of the equivalent circuit parameters. The tests run in variable load conditions in a usual interval of operation of the motor, in terms of slip, show that the equivalent circuit does not represent quit well the behavior of the machine when it works out of the rated point, because the parameters do not represent it in those operating condition any more. The magnitudes obtained in the variable load test are conditioned and used in the training of an artificial neural network Multi-Layer Perceptron (MLP). For the validation of the results, the neural model and experimental tests of the induction motor results are compared.
Mestrado
Automação
Mestre em Engenharia Elétrica

The transportation community is exploring how new "big'' databases constructed by companies or public administrative agencies can be used to better understand travelers' behaviors and better predict travelers' responses to various transportation policies. This thesis explores how a large targeted marketing database containing information about individuals’ socio-demographic characteristics, current residence attributes, and previous residential locations can be used to investigate research questions related to individuals' transportation preferences and the built environment. The first study examines how household vehicle ownership may be shaped by, or inferred from, previous behavior. Results show that individuals who have previously lived in dense ZIP codes or ZIP codes with more non-automobile commuting options are more likely to own fewer vehicles, all else equal. The second study uses autoregressive models that control for spatial dependence, correlation, and endogeneity to investigate whether investments in public transit infrastructure are associated with higher home values. Results show that willingness-to-pay estimates obtained from the general spatial Durbin model are less certain than comparable estimates obtained through ordinary least squares. The final study develops an empirical framework to examine a housing market's resilience to price volatility as a function of transportation accessibility. Two key modeling frameworks are considered. The first uses a spatial autoregressive model to investigate the relationship between a home's value, appreciation, and price stability while controlling for endogenous missing regressors. The second uses a latent class model that considers all these attributes simultaneously, but cannot control for endogeneity.

PAE
O transporte rodoviário é um dos modais mais utilizados para o transporte de passageiros principalmente devido à sua acessibilidade, tanto no contexto nacional como mundial. As organizações buscam métodos e sistemas cada vez mais eficientes para atender de forma satisfatória seus clientes, oferencendo serviços de qualidade e confiáveis. O presente trabalho teve como objetivo propor um modelo matemático de veículos para a minimização de ônibus, evitando ao máximo o desbalanceamento da rede de transporte. O modelo buscou abranger as restrições inerentes aos fluxos em rodoviárias de modo a cobrir todas as viagens de uma tabela de horários já definida, alocando apenas um veículo para cada viagem e aproveitando o mesmo para outras viagens quando possível, atendendo à função objetivo. O modelo foi aplicado à uma empresa de transportes de passageiros com quatro cidades envolvidas e 105 viagens em um tempo de 24 horas. As variáveis e restrições foram identificadas a partir de entrevistas semi -estruturadas e visitas à organização, observando roteiros de viagens a serem cumpridos por dois modelos de ônibus: metropolitanos e articulados. Neste contexto, dois modelos distintos foram montados de forma a atender as viagens, sendo analisadas separadamente. Os resultados definiram um total necessário de 12 ônibus metropolitanos e 4 ônibus articulados para atender às exigências da empresa e da tabela viagens, sendo que todas as restrições foram satisfeitas, detectando -se a necessidade da realização de seis viagens mortas (sem passageiros) para suprir a necessidade de d emanda de ônibus para cada tabela de viagens diárias. Com os resultados, fez-se uma comparação com a programação que a empresa necessitava para cumprir a tabela de viagens, identificando a diminuição de um ônibus metropolitano, sem deixar de cumprir a restrição de autonomia de cada ônibus poder rodar 400 quilômetros sem necessidade de reabastecimento. Algumas melhorias foram identificadasanalisando-se a rede resultante, sendo discutidas ao final da pesquisa e norteando novas intenções de trabalhos futuros.
Road transport is one of the most used modes for passenger transport mainly due to its affordability, in the national and global context. Organizations seek for a better methods and systems to meet the satisfaction of its customers, offering quality and reliable services. This research aimed to propose a vehicle scheduling model to minimize buses, avoiding the transmission system imbalance. The model tried to cover the restrictions inherent about the vehicle flow on bus stations to cover all travel in a given timetable, allocating only one vehicle for each trip and using it for other trips when it is possible, given the objective function. The model was applied to a passenger transportation company with four cities involved and 105 trips in a time of 24 hours. The variables and constraints were identified from semi-structured interviews and visits at organization, identifying the tours to be met by two bus models: articulated and metropolitan. In this context, two different models were fitted to meet the travel , and analyzed separately. The results defined a required total of twelve metropolitan buses and four articulated buses to meet the company and timetable requirements and all constraints were satisfied, detecting the need of perform six deadheading trips (without passengers) to meet the need for demand bus for each table in the daily trips. With the results, a comparison was made between the old programming system adopted by the organization to fulfill the timetable, identifying the reduction of a metropolitan bus, while fulfilling the constraint of autonomy that each bus can travel 400 kilometers. Some improvements were identified by observing the resulting route network, being discussed at the end of the research and guiding for a new works in the future.

En la presente tesis se desarrolla un sistema de control de la mezcla de aire y gasolina en un motor alternativo de cuatro tiempos, basado en redes neuronales. Para que el catalizador logre un grado de depuración aceptable con todos los gases contaminantes simultáneamente, debe mantenerse la proporción de aire y combustible, dentro de una banda muy estrecha. En estado estacionario, este objetivo se cumple sin demasiados problemas, pero el funcionamiento habitual de un vehículo es en régimen muy transitorio, donde los sistemas convencionales no logran evitar desviaciones importantes del punto de consigna. Estos se basan además en una gran cantidad de tablas estáticas, que deben calibrarse de forma experimental, lo cual es bastante costoso tanto en tiempo como en dinero.
Para evitar estos problemas se ha diseñado un modelo matemático de un motor, que comprende todo el proceso de formación de la mezcla, sensores, la generación del par motor y la dinámica del vehículo. La realización final se ha realizado en el lenguaje de simulación MatLab/Simulink®. Los datos requeridos son fácilmente obtenibles bien por metrología, bien de forma experimental. Se ha validado con un motor SEAT de 1,6 l y 74 kW.
Como primer paso se ha aplicado una estrategia de control convencional bastante simple, consistente en un controlador feedforward estático, más un controlador feedback de tipo PI ó PID. Esto ha permitido el estudio de las principales características del motor desde el punto de vista de control.
Con los resultados obtenidos se ha diseñado un observador basado en una red neuronal, que elimine los retardos puros del sistema y que pueda ser utilizado para cerrar el lazo de control. Primero se ha usado una red feedforward, pero vistos los malos resultados, se ha desarrollado una red neuronal recurrente a partir de la red de Elman, que se ha modificado convenientemente para adaptarla a las dificultades propias del problema. El algoritmo de entrenamiento utilizado se basa en el de retropropagación clásico, y modifica no sólo los pesos entre capas, sino también los correspondientes a las neuronas de contexto, las cuales permiten memorizar estados internos. La principal mejora consiste en separar las neuronas de contexto en tantos grupos como entradas tiene la red, y entrenarlos por separado, de modo que cada grupo se adapte a la dinámica particular de la entrada a la que va asociado. Se muestra mediante simulación el comportamiento del conjunto motor más observador en lazo cerrado, y se compara con el esquema convencional. Se prueba asimismo la robustez del sistema frente a distintas consignas, ruido en la planta y defectos de sintonía.
In the present thesis a control system for the air-fuel mixture in a reciprocating four-stroke engine is developed, based on neural networks. The air-fuel ratio has to be kept within a very narrow window so that the catalyst achieves an acceptable degree of purification simultaneously with all the polluting gases. In steady state, this goal can be fulfilled without difficulties, but the usual operation of a vehicle is in a very transient state, where the conventional systems are not able to avoid important excursions from the set point. They also rely on a great number of look-up tables, which have to be tuned experimentally, thus with an enormous investment of money and time.
To avoid those problems a mathematical model of an engine has been designed, in such a way that it comprises of the whole mixture formation process, sensors, the torque generation and the vehicle dynamics. Finally it has been implemented in the simulation language MatLab/Simulink®. The required data is easily available both from metrology, and experimental work. It has been validated with a 1,6 litre 74 kW SEAT engine.
As a first step, a quite simple conventional control strategy has been applied, consisting of a static feedforward controller, and a PI or PID feedback controller. This has permitted studying the main features of the engine from the control point of view.
With the obtained results an observer based on a neural network has been designed, which eliminates the delays of the system and that can be used to close the control loop. First a feedforward network has been used, but due to the bad results, a recurrent neural network has been developed starting from the Elman network, which has been properly modified in order to adapt it to the characteristic difficulties of the problem. The training algorithm used is based on that of classical backpropagation, and it modifies not only the weights interconnecting different layers, but also those corresponding to the context neurons, which allow the memorising of internal states. The main improvement consists in separating the context neurons in as many groups as the network has inputs, and to train them separately, so that each group adapts to the particular dynamics of the input with which it is associated. The behaviour of the engine plus the observer in closed loop is shown by means of simulation, and is compared with the conventional scheme. It is proven the robustness of system response to different set points, noise in the plant and tuning defects.

Краснов, Р. В. Підвищення надійності електродвигунів компресорів електропоїздів постійного струму : авт. дис. к. т. н.: 05.22.09 / Р. В. Краснов ; Дніпропетр. нац. ун-т залізн. трансп. ім. акад. В. Лазаряна. - Д., 2012.
UK: АНОТАЦІЯ Дисертація присвячена підвищенню надійності електродвигунів компресорів електрорухомого складу шляхом забезпечення нормованих значень нагрівання ізоляції в реальних умовах експлуатації і удосконалення системи захисту якірно-го кола електродвигуна від надмірних струмів, які можуть викликати нагрівання ізоляції більше допустимих значень. Пропонується, на відміну від попередніх досліджень, процес пуску розгля-дати таким, що складається з двох етапів. В роботі проведено дослідження по визначенню значень струмів на кожному з цих етапів і впливу тривалості етапів на нагрівання ізоляції. Для теоретичних досліджень автором запропонована ма-тематична модель, адекватність якої підтверджена тим, що різниця між параме-трами, які отримані з допомогою цієї моделі та експериментально (шляхом ос-цилографування), знаходиться в межах до 5 %. Отримані залежності тривалості першого та другого етапів пуску, а також значень струму якоря на цих етапах при різних умовах: різних значеннях на-пруги в контактній мережі, температурі навколишнього середовища, тиску в пневмомережі та значенням моменту на валу двигуна після закінчення пуску. Експериментально та теоретично доведено, що в окремих випадках захист електродвигуна з допомогою теплового реле ТРВ-8,5 не забезпечується. З допомогою теплової моделі досліджені процеси нагрівання ізоляції обмо-тки якоря під час пуску та подальшої роботи електродвигуна компресора при різних можливих режимах та умовах. Результати досліджень, що отримані за допомогою математичної та тепло-вої моделей використані для програмування нового пристрою захисту електро-двигуна компресора від перенавантажень. Запропонований пристрій захисту безвідмовно протягом декількох років працює на реальному електропоїзді. RU: АННОТАЦИЯ Диссертация посвящена повышению надежности электродвигателей ком-прессоров электроподвижного состава путём обеспечения рациональных значе-ний нагревания изоляции в реальных условиях эксплуатации и усовершенство-вания системы защиты якорной цепи электродвигателя от чрезмерных токов, которые могут вызвать нагревание изоляции больше нормированных значений. Проведен анализ выхода из строя электродвигателей компрессоров на электропоездах постоянного тока серии ЭР-1, 2 по причине пробоя изоляции обмоток якоря и полюсов. Как показывает обзор литературы по теме диссерта-ционной работы, проведенный автором, до теперешнего времени влияние про-цесса пуска на нагревание изоляции обмоток якоря и полюсов в полной мере не исследовано, особенно это относится к рассмотрению процесса пуска как тако-вого. Автор, в отличие от предыдущих исследований, предлагает процесс пуска рассматривать состоящим из двух этапов. Первый этап пуска – напряжение на электродвигатель подано, а его вал не вращается в течении некоторого времени. Второй этап – вал электродвигателя трогается и начинает набирать обороты до установившегося значения угловой скорости. В работе проведено исследование по определению значений токов на каж-дом из этих этапов и влиянию длительности этапов на нагревание изоляции. Для этого вначале проведено экспериментальное исследование процессов пуска электродвигателя компрессора на реальном электропоезде путем осциллогра-фирования этого процесса с помощью современного измерительного комплекса с исользованием датчиков напряжения (тип LEM CV4-6000) и тока (тип LEM LT100-8/SP30). Для теоретических исследований автором предложена математическая мо-дель, адекватность которой подтверждена тем, что расхождение параметров, полученных путем осциллографирования и с помощью математической модели находится в пределах до 5 %. Для получения численных значений при исследо-вании процессов пуска на указанных двух этапах с помощью математической модели используется метод Рунге-Кутта-Фелберга. Получены зависимости тока якоря на первом и втором этапах пуска при разных условиях: разных значениях напряжения в контактной сети, температу-ры окружающей среды, давления в пневматической сети. Получены также зависимости длительности первого и второго этапов пус-ка от указанных условий. Установлено, что в реальных условиях эксплуатации длительность первого этапа пуска составляет от 19 до 74 мс, однако, несмотря на относительно малое значение, токи, возникающие на первом этапе, существенно влияют на эффек-тивное значение тока за всё время пуска и соответственно существенно влияют на нагревание изоляции. Так увеличение длительности первого этапа пуска до 30 % по сравнению с пуском электродвигателя при номинальном напряжении в контактной сети и нормальных внешних условиях увеличивает эквивалентный ток за всё время пуска до 9,1 %. Установлены зависимости между длительностью первого этапа и значени-ем момента на валу двигателя после окончания пуска. Экспериментальные и теоретические исследования показали, что в отдель-ных случаях защита электродвигателя с помощью теплового реле ТРВ-8,5 не обеспечивается. С помощью тепловой модели исследованы процессы нагревания изоляции обмотки якоря во время пуска и дальнейшей работы электродвигателя компрес-сора при номинальной нагрузке на его валу и при бо́льших нагрузках, что воз-можно при реальных условиях эксплуатации. Доказано, что в эксплуатации возможны режимы работы, при которых из – за перегрева изоляции срок её службы снижается до 4,8 лет, что соответствует статистическим данным. Установлена зависимость времени, за которое протекание тока определён-ного значения приводит к максимально допустимому нормированному нагре-ванию изоляции обмоток якоря от этого тока при наиболее неблагоприятной температуре окружающей среды – +40 ˚С. Эта зависимость использована для программирования предложенного автором нового устройства защиты якорных цепей на базе современных микроконтроллеров. Предложенное автором устройство защиты от перегрузок электродвигателя компрессора безотказно в течение нескольких лет работает на реальном элек-тропоезде. Проведено технико – экономическое обоснование внедрения результатов работы. EN: ABSTRACT The dissertation paper is devoted to improving the reliability of electromotor compressors of rolling stock by providing normative values of heat insulation under real operating conditions and improve protection anchor range from excessive elec-tric currents that can cause heat insulation more acceptable values. Proposed, in contrast to previous studies, the process of starting to consider as being composed of two stages. In this paper a study to determine the values of currents at each of these phases and the influence of the duration of the stages of heat insulation. For theoretical investigations of the author to proposed mathematical model, whose adequacy is confirmed that the difference between the parameters ob-tained using this model and experimentally (by oscillography) is within 5%. Dependences of the duration of the first and second stages of start-up and current values of anchor on these stages under different conditions: different values of voltage in the contact system, ambient temperature, air pressure in the pneumatic system and the value of the moment on the shaft of the electric motor after starting. Experimentally and theoretically shown that in some cases protection of the electric motor with thermal relay ТРВ-8, 5 is not provided. With the thermal model the processes of heat insulation of armature winding during start-up and subsequent operation of the electric motor of compressor for various possible modes and conditions are investigated. The research results, obtained by using mathematical and thermal models, used for programming a new protection device of electric motor compressor from overload. The proposed protection device flawlessly for several years working on real electric rolling stock.

Investigations into the combustion of alternative fuels in compression-ignition engines in South Africa have underlined the inadequacies of existing zero-dimensional combustion models. The major aspect of concern in these models was the computation of heat transfer which had been singled out by a number of researchers as the leading cause of inaccuracies in heat release computations. The main objective of this research was to develop a combustion model that was less empirically based than the existing zerodimensional models for use in evaluating the combustion and resulting thermal stresses generated by alternative fuels. in diesel engines. Particular attention was paid to the development of a spatial and temporal model of convective heat transfer that was based on gas flow characteristics and to the introduction of a radiation heat transfer model that made use of fuel properties and fuel-air ratio. The combustion process was divided into two zones representing burnt and unburnt constituents and the resulting temperatures in each zone were used in the calculations of convective and radiative heat transfer. The complete model was formulated in such a way that it could be applied with the aid of a micro-computer. Calibration and verification of the gas flow sub-models which involved the squish, swirl and turbulence components necessitated the use of published data. Good agreement for the squish and swirl components was obtained between the present model and the experimental data from three engines, two with a bowl-in-piston and the other with a flat piston. These gas flow components dominated the gas velocities in the combustion chamber and provided a reliable foundation for the calculation of convective heat transfer. In spite of the well documented difficulties of characterising turbulence, after calibration the model generated turbulence levels with acceptable trends and magnitudes. Tests were carried out on a naturally aspirated ADE 236 engine involving the measurement of cylinder pressure and heat flux at a single point. Motored engine data were used to verify the convective heat transfer rates and to ascertain the effects of soot deposition on the heat flux probe. Close correlation between predicted and measured heat flux was achieved after accounting for the effects of chamber geometry at the probe site. Soot deposition on the probe caused a significant attenuation of the heat flux within a short period of the engine running under fired conditions. The results from fired engine tests showed that the two zone combustion model was providing plausible trends in the burnt and unburnt zone temperatures and that the model generated combined heat transfer rates which were credible not only on a global basis but also in terms of point predictions in the combustion chamber. The results also highlighted the considerable variation in heat transfer that could occur from one point in the chamber to another. Such variations added considerable weight to the objective of moving away from a zero-dimensional model to a quasi-dimensional type where predictions could be made on a more localised rather than global basis. It was concluded that the model was a definite improvement over zero-dimensional models and competed favourably with existing quasi-dimensional models with advantages in both simplicity and accuracy.
Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1989.

Thesis (D.Tech.: Mechanical Engineering)-Dept. of Mechanical Engineering, Durban Institute of Technology, 2005 xxi, 265 leaves
The aim of this study is to develop, using neural networks, a model to aid the performance monitoring of operational diesel engines in industrial settings. Feed-forward and modular neural network-based models are created for the prediction of the specific fuel consumption on any normally aspirated direct injection four-stroke diesel engine. The predictive capability of each model is compared to that of a published quadratic method. Since engine performance maps are difficult and time consuming to develop, there is a general scarcity of these maps, thereby limiting the effectiveness of any engine monitoring program that aims to manage the fuel consumption of an operational engine. Current methods applied for engine consumption prediction are either too complex or fail to account for specific engine characteristics that could make engine fuel consumption monitoring simple and general in application. This study addresses these issues by providing a neural network-based predictive model that requires two measured operational parameters: the engine speed and torque, and five known engine parameters. The five parameters are: rated power, rated and minimum specific fuel consumption bore and stroke. The neural networks are trained using the performance maps of eight commercially available diesel engines, with one entire map being held out of sample for assessment of model generalisation performance and application validation. The model inputs are defined using the domain expertise approach to neural network input specification. This approach requires a thorough review of the operational and design parameters affecting engine fuel consumption performance and the development of specific parameters that both scale and normalize engine performance for comparative purposes. Network architecture and learning rate parameters are optimized using a genetic algorithm-based global search method together with a locally adaptive learning algorithm for weight optimization. Network training errors are statistically verified and the neural network test responses are validation tested using both white and black box validation principles. The validation tests are constructed to enable assessment of the confidence that can be associated with the model for its intended purpose. Comparison of the modular network with the feed-forward network indicates that they learn the underlying function differently, with the modular network displaying improved generalisation on the test data set. Both networks demonstrate improved predictive performance over the published quadratic method. The modular network is the only model accepted as verified and validated for application implementation. The significance of this work is that fuel consumption monitoring can be effectively applied to operational diesel engines using a neural network-based model, the consequence of which is improved long term energy efficiency. Further, a methodology is demonstrated for the development and validation testing of modular neural networks for diesel engine performance prediction.

The evaluation of the stray-load losses in induction motors is an important part of the efficiency estimation process. However, stray-load losses have several components which are difficult to isolate and calculate. Consequently, they are more readily determined by test. IEEE Standard 112-1996 recommends the testing of electrical induction motors at a rated mode (nominal frequency and voltage). Thus, the IEEE 112 test does not provide us with information about the motor efficiency and stray load losses at supply frequencies which are different from the rated value. At the same time, this information is very important for the correct evaluation of the efficiency of the motors, controlled by the Adjustable Speed Drives (ASDs). This work is dedicated to the investigation of the relation between the motor efficiency, stray-load loss and the supply frequency. The voltages to be applied for the specific values of the frequency were chosen to keep the value of the relative slip constant at 100% load torque. After performing the series of load tests, it was found that the SLL is proportional to the square of the load and the dependency between SLL and frequency for different load levels is close to linear. The comparison of the efficiency values obtained by the methods IEEE 112 B and 112 E has shown a good correlation of the results.
Graduation date: 2004

A Research Report submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, in partial fulfilment of the requirements for the degree of Master of Science in Engineering (MSc 50/50) School of Chemical and Metallurgical Engineering, Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, South Africa November, 2016
The global focus on cleaner air has seen sulphur removal processes’ gaining popularity and adsorptive desulphurisation has been identified as an effective alternative. Adsorptive desulphurisation was used to simulate and evaluate the performance of the polymer supported imidation agent (Sodium N-chloro-polystyrene sulphonamide) as an adsorbent in the desulphurisation of diesel fuel. This study involved the development of a mathematical model for the adsorption process of sulphur on the polymer supported imidation agent, based on the mass balance on a continuous fixed bed column and pseudo second order kinetics. The developed model was solved using numerical methods, and the simulation of the process carried out varying different parameters; the inlet sulphur concentration, the adsorption column bed height and the particle size (radius) of the adsorbent. The simulation showed that the adsorption capacity of the studied adsorbent increased with increase in the inlet sulphur concentration; an increase in the adsorption bed height and a decrease in the adsorbent particle size. Validation of the simulation done was carried out by comparing the simulation data with experimental data. The proposed model fit experimental data and can be used to predict the inlet concentration conditions, bed height and particle size of the adsorbent. The overall research enhances the understanding of the adsorptive desulphurisation of diesel fuel using the polymer supported imidation agent and the mathematical modelling of the process.
MT2017

In this dissertation, a group of vehicle dynamics simulation tools is developed with two primary goals: to accurately represent vehicle behavior and to provide insight that improves the understanding of vehicle performance. Three tools are developed that focus on tire modeling, vehicle modeling and lap time simulation. Tire modeling is based on Nondimensional Tire Theory, which is extended to provide a flexible model structure that allows arbitrary inputs to be included. For example, rim width is incorporated as a continuous variable in addition to vertical load, inclination angle and inflation pressure. Model order is determined statistically and only significant effects are included. The fitting process is shown to provide satisfactory fits while fit parameters clearly demonstrate characteristic behavior of the tire. To represent the behavior of a complete vehicle, a Nondimensional Tire Model is used, along with a three degree of freedom vehicle model, to create Milliken Moment Diagrams (MMD) at different speeds, longitudinal accelerations, and under various yaw rate conditions. In addition to the normal utility of MMDs for understanding vehicle performance, they are used to develop Limit Acceleration Surfaces that represent the longitudinal, lateral and yaw acceleration limits of the vehicle. Quasi-transient lap time simulation is developed that simulates the performance of a vehicle on a predetermined path based on the Limit Acceleration Surfaces described above. The method improves on the quasi-static simulation method by representing yaw dynamics and indicating the vehicle's stability and controllability over the lap. These improvements are accomplished while maintaining the simplicity and computational efficiency of the two degree of freedom method.
Graduation date: 2013

M. Tech. : Mathematical Technology.
Attempts will be made to extend the elementary quarter-mass models (for instance Gillepse, 1992, [5]; Kiecke & Nielsen, 2000, [6] and Singiresu, 2004, [7]) of a motor vehicle suspension system to include the radial vibrations of a rubber tyre in the model. Tangential vibrations of the tyre surface were investigated by Bekker (2009, [8]) and the possible incorporation of such vibrations into a suspension model invites the possibility of future study.

by Meng Xiaoning.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2006.
Includes bibliographical references (leaves 99-106).
Abstracts in English and Chinese.
Chapter 1 --- Introduction --- p.1
Chapter 1.1 --- Motivation --- p.1
Chapter 1.2 --- Overview --- p.2
Chapter 1.3 --- Organization of the Thesis --- p.3
Chapter 2 --- Related Research --- p.6
Chapter 2.1 --- Information Technology in Cars --- p.6
Chapter 2.2 --- Anti-theft Protection --- p.8
Chapter 2.3 --- Learning Human Behaviors --- p.10
Chapter 2.4 --- Neural Network Learning --- p.11
Chapter 3 --- Experimental Design --- p.14
Chapter 3.1 --- Overview --- p.14
Chapter 3.2 --- Driving Simulation Subsystem --- p.14
Chapter 3.3 --- Data Sensing and Capturing Subsystem --- p.15
Chapter 3.4 --- Data Analysis Subsystem --- p.17
Chapter 4 --- Data Preprocessing for Feature Selection --- p.23
Chapter 4.1 --- Introduction --- p.23
Chapter 4.2 --- Fast Fourier Transform --- p.23
Chapter 4.3 --- Principal Component Analysis --- p.24
Chapter 4.4 --- Independent Component Analysis --- p.26
Chapter 5 --- Classification via Support Vector Machine --- p.28
Chapter 5.1 --- Introduction --- p.28
Chapter 5.1.1 --- Why Using Support Vector Machine --- p.28
Chapter 5.1.2 --- Mathematic Description --- p.29
Chapter 5.2 --- Problem Formulation --- p.31
Chapter 5.3 --- Approach --- p.31
Chapter 5.4 --- Experimental Results --- p.34
Chapter 5.4.1 --- Preprocess Data Analysis --- p.34
Chapter 5.4.2 --- Models Design --- p.37
Chapter 5.5 --- Discussion --- p.44
Chapter 6 --- Evaluation via Hidden Markov Model --- p.47
Chapter 6.1 --- Introduction --- p.47
Chapter 6.1.1 --- Why Using Hidden Markov Model --- p.48
Chapter 6.1.2 --- Mathematic Description --- p.50
Chapter 6.2 --- Problem Formulation --- p.51
Chapter 6.3 --- Approach --- p.53
Chapter 6.4 --- Experimental Results --- p.56
Chapter 6.4.1 --- Model-to-model Measure --- p.56
Chapter 6.4.2 --- Human-to-model Measure --- p.63
Chapter 6.4.3 --- Parameters Optimization --- p.66
Chapter 6.5 --- Discussion --- p.69
Chapter 7 --- System Design and Implementation --- p.71
Chapter 7.1 --- Introduction --- p.71
Chapter 7.2 --- Hardware --- p.72
Chapter 7.3 --- Software --- p.78
Chapter 7.4 --- System Demonstration --- p.80
Chapter 8 --- Conclusion and Future Work --- p.82
Chapter 8.1 --- Contributions --- p.82
Chapter 8.2 --- Future Work --- p.84
Chapter A --- Hidden Markov Model Training --- p.87
Chapter A.1 --- Forward-backward Algorithm --- p.87
Chapter A.2 --- Baum-Welch Algorithm --- p.87
Chapter B --- Human Driving Behavior Data --- p.90
Chapter C --- Publications Resulted from the Study --- p.98

As theoretical neuroscience has grown as a field, machine learning techniques have played an increasingly important role in the development and evaluation of theories of neural computation. Today, machine learning is used in a variety of neuroscientific contexts from statistical inference to neural network training to normative modeling. This dissertation introduces machine learning techniques for use across the various domains of theoretical neuroscience, and the application of these techniques to build theories of neural circuits. First, we introduce a variety of optimization techniques for normative modeling of neural activity, which were used to evaluate theories of primary motor cortex (M1) and supplementary motor area (SMA). Specifically, neural responses during a cycling task performed by monkeys displayed distinctive dynamical geometries, which motivated hypotheses of how these geometries conferred computational properties necessary for the robust production of cyclic movements. By using normative optimization techniques to predict neural responses encoding muscle activity while ascribing to an “untangled” geometry, we found that minimal tangling was an accurate model of M1. Analyses with trajectory constrained RNNs showed that such an organization of M1 neural activity confers noise robustness, and that minimally “divergent” trajectories in SMA enable the tracking of contextual factors. In the remainder of the dissertation, we focus on the introduction and application of deep generative modeling techniques for theoretical neuroscience. Specifically, both techniques employ recent advancements in approaches to deep generative modeling -- normalizing flows -- to capture complex parametric structure in neural models. The first technique, which is designed for statistical generative models, enables look-up inference in intractable exponential family models. The efficiency of this technique is demonstrated by inferring neural firing rates in a log-gaussian poisson model of spiking responses to drift gratings in primary visual cortex. The second technique is designed for statistical inference in mechanistic models, where the inferred parameter distribution is constrained to produce emergent properties of computation. Once fit, the deep generative model confers analytic tools for quantifying the parametric structure giving rise to emergent properties. This technique was used for novel scientific insight into the nature of neuron-type variability in primary visual cortex and of distinct connectivity regimes of rapid task switching in superior colliculus.

CD-ROM containing source codes of the numerical scheme (appendix A) is attached to back cover.
Includes bibliographical references (leaves 459-472).
System requirements for accompanying CD-ROM: Macintosh or IBM compatible computer. Other requirements: Adobe Acrobat Reader.
xxxvi, 473 leaves ; ill. ; 30 cm. + 1 computer optical disk (4 3/4 in.)
Title page, contents and abstract only. The complete thesis in print form is available from the University Library.
A numerical scheme, based on discrete-vortex and surface-vorticity boundary-integral methods, has been developed for stimulating time dependent, two-dimensional, viscous flow over arbitary arrays of solid bodies of arbitary cross-section
Thesis (Ph.D.)--Adelaide University, Dept. of Mechanical Engineering, 2001

CD-ROM containing source codes of the numerical scheme (appendix A) is attached to back cover.
Includes bibliographical references (leaves 459-472).
System requirements for accompanying CD-ROM: Macintosh or IBM compatible computer. Other requirements: Adobe Acrobat Reader.
xxxvi, 473 leaves ; ill. ; 30 cm. + 1 computer optical disk (4 3/4 in.)
A numerical scheme, based on discrete-vortex and surface-vorticity boundary-integral methods, has been developed for stimulating time dependent, two-dimensional, viscous flow over arbitary arrays of solid bodies of arbitary cross-section
Thesis (Ph.D.) -- University of Adelaide, Dept. of Mechanical Engineering, 2001

The United States Environmental Protection Agency has estimated that over 111 million people reside in areas that exceed the National Ambient Air Quality Standards for ozone. One major source of the chemical precursors (nitrogen dioxides and volatile organic compounds (VOCs)) for ozone are motor vehicles. The overall goal of this research is to improve the knowledge base related to VOC refueling and evaporative emissions from motor vehicles. Refueling, running loss, hot soak, and diurnal loss total and speciated VOC emissions were investigated. A total of 12 uncontrolled refueling events were completed and involved the determination of volumetric flow rates of gasoline vapor during refueling, as well as total and speciated VOC concentrations. Total VOC emissions were compared with two commonly used algorithms. Speciated VOC vapor profiles were compared with two published gasoline vapor profiles and theoretical predictions based on knowledge of liquid composition and environmental conditions. An evaluation of refueling emissions impacts on ozone formation potentials using MIR was completed and results were compared against speciated emissions and MOBILE-based total VOC emissions estimates coupled with a default speciation profile. Refueling VOC emissions and resultant ozone formation potential may be underestimated in existing emission inventories, particularly during the summer ozone season, A model was developed to predict the speciation of VOCs associated with evaporative emissions from motor vehicles. Model-predicted speciation profiles were evaluated using SHED studies. Running loss, hot soak and diurnal emissions were included in each test. Total VOC emissions measured during each test were compared against MOBILE6 predicted emissions. An evaluation of evaporative emissions impacts on ozone formation potentials using MIR was completed, comparing measured and predicted emissions. The measured:predicted speciation results ranged between 0.93 and 1.11 and had an average value of 1.02. For the conditions tested, MOBILE6 underestimated evaporative emissions in 20 of 24 comparisons. MOBILE6-based ozone formation potentials may be underestimated.

Indiana University-Purdue University Indianapolis (IUPUI)
Lithium-ion battery is one kind of rechargeable battery, and also renewable, sustainable and portable. With the merits of high density, slow loss of charge when spare and no memory effect, lithium-ion battery is widely used in portable electronics and hybrid vehicles. Apart from its advantages, safety is a major concern for Lithium-ion batteries due to devastating incidents with laptop and cell phone batteries. Overcharge and over-discharge are two of the most common electrical abuses a lithium-ion battery suffers. In this thesis, a fuzzy-rule-based system is proposed to detect the over-charge and over-discharge failure in early time. The preliminary results for the failure signatures of overcharged and over-discharged lithium-ion are listed based on the experimental results under both room temperature and high temperature. A fuzzy-rule-based model utilizing these failure signatures is developed and validated. For over-charge case, the abnormal increase of the surface temperature and decrease of the voltage are captured. While for over discharge case, unusual temperature increase during overcharge phases and abnormal current decrease during overcharge phases are obtained. The inference engine for fuzzy-rule-based system is designed based on these failure signatures. An early warning signal will be given by this algorithm before the failure occurs. This failure detection and early warning system is verified to be effective through experimental validation. In the validation test, the proposed methods are successfully implemented in a real-time system for failure detection and early warning. The result of validation is compatible with the design expectation. Finally an accurate failure detection and early warning system is built and tested successfully.

Indiana University-Purdue University Indianapolis (IUPUI)
We present a novel, variational and statistical approach for model-based segmentation. Our model generalizes the Chan-Vese model, proposed for concurrent segmentation of multiple objects embedded in the same image domain. We also propose a novel shape descriptor, namely the Multi-Compartment Distance Functions or mcdf. Our proposed framework for segmentation is two-fold: first, several training samples distributed across various classes are registered onto a common frame of reference; then, we use a variational method similar to Active Shape Models (or ASMs) to generate an average shape model and hence use the latter to partition new images. The key advantages of such a framework is: (i) landmark-free automated shape training; (ii) strict shape constrained model to fit test data. Our model can naturally deal with shapes of arbitrary dimension and topology(closed/open curves). We term our model Active Geometric Model, since it focuses on segmentation of geometric shapes. We demonstrate the power of the proposed framework in two important medical applications: one for morphology estimation of 3D Motor Neuron compartments, another for thickness estimation of Henle's Fiber Layer in the retina. We also compare the qualitative and quantitative performance of our method with that of several other state-of-the-art segmentation methods.